Radiant energy – Photocells; circuits and apparatus – Photocell controlled circuit
Reexamination Certificate
1999-06-18
2001-06-19
Le, Que T. (Department: 2878)
Radiant energy
Photocells; circuits and apparatus
Photocell controlled circuit
C257S444000
Reexamination Certificate
active
06248992
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to photoconductive semiconductor switches. Photoconductive semiconductor switches (PCSS) are under investigation for use as short duration, high current semiconductor devices. A PCSS in the lateral configuration is shown in FIG.
1
A. More particularly, this invention relates to such switches implemented in gallium arsenide that have electric fields in excess of the lock-on voltage impressed across their terminals at the time laser light is directed upon them to trigger the mobilization of the charge carriers in the switch.
PCSS based on the lock-on effect have previously been fabricated in GaAs. The switches discussed herein are GaAs switches laid out in a lateral or vertical configuration as shown in
FIGS. 1A and 1B
respectively. At electric fields below about 4 kV/cm, GaAs switches are activated by the creation of, at most, one electron hole pair per photon absorbed. This linear mode demands high laser power, and, after the laser light is extinguished, the carrier density decays exponentially in 1-10 ns. At higher electric fields these switches behave very differently. On triggering, the high field induces carrier multiplication so that the amount of light required is reduced by as much as five orders of magnitude. We have used trigger energies as low as 180 nJ to deliver 48 MW in a 30-50 ohm system and as low as 2 nJ from a vertical cavity surface emitting laser (VCSEL). In the ‘on’ state the field across the switch stabilizes to a constant called the lock-on field. The switch current is circuit-limited provided the circuit maintains the lock-on field. As the initial (prior to triggering) field increases, the switch risetime decreases and the trigger energy is reduced. During high gain switching the switches emit bandgap radiation due to carrier recombination. When this radiation is imaged, filaments are observed, even if the triggering radiation is uniform. The filaments can have densities of MA/cm
2
and diameters of 15-300 &mgr;m.
These switches can be used for pulsed power applications as diverse as low impedance, high current pulsers and high impedance, low current Pockels cell or Q switch drivers. Advances in this technology offer improvements over alternative switching schemes, i.e. 100 ps risetime, kilohertz (continuous) and megahertz (burst) repetition rates; they are scalable or stackable to hundreds of kilovolts and tens of kiloamps, useful for optical control and isolation, and have inherent solid state reliability.
However, existing PCSS switches are limited to use in applications where the switches are to be used for a limited number of pulses. The current filamentation causes damage to the switch at the point where the current enters or exits the insulating, un-doped, region of the switch (the intrinsic region). What is needed is an improved GaAs PCSS with higher longevity.
High current density in the PCSS leads to localized heating during operation and consequent reliability problems. Although PCSS have been demonstrated to operate at currents in excess of 1000 amperes and switch voltages in excess of 100 kV, no practical devices have demonstrated sufficient reliability to be used in a real application. Depending on the operating current level and the type of charging circuit used, the switch life has not exceeded 2×10
5
pulses for a standard test condition with a pin structure and strong light trigger. The next section will address different triggering methods. Current levels in excess of 20 A do not even achieve this lifetime, with the life dropping dramatically as the current is increased. For example, a 1000 A current level has been achieved for only 2 pulses. When the switch degrades, the damage is observed primarily at one or both of the metal contacts.
A number of approaches to alleviate the damage to PCSS while maintaining its high current and short pulse duration have been attempted. It has been shown that diffuse triggering near a given contact leads to multiple filaments near that contact. Instead of triggering the filament with a single localized pulse of light, the trigger light is spatially spread out over a larger area adjacent to the contact. When the filament is initiated, the filament image consists of multiple filaments near the source of the triggering light. By the use of this technique, the damage near the contact of the diffuse triggered region is greatly lessened. For example, in a lateral PCSS with a p-type ohmic contact for the anode and an n-type contact for the cathode, the p-type contact erodes first with conventional localized triggering. However, a diffused light trigger near the p-contact results in multiple filaments near the p-contact and results in a dramatic increase in the useful life of the PCSS prior to the onset of damage. Great improvements in longevity were obtained at about 10 A, where millions of switching actions were obtained with no damage. When damage is observed, the n-type contact is more heavily damaged. The use of dual optical fibers to diffusely trigger both sides of the PCSS has proven beneficial, but still does not lead to sufficiently long operation of the PCSS at useful current levels. For example, when switching a voltage of 9,000 Volts, discharging a current of 70 A for 3.5 ns, a diffuse-triggered PCSS only lasted 650,000 pulses and had considerable damage.
Ion implantation or diffusion has also been suggested as a method of reducing the current crowding (or current pinching or increasing the transfer length) at the edge of the contact to increase PCSS lifetime. However, neither method specifically articulated any detail other than that noted above.
Significant effort has addressed the longevity problem by improving the quality of the contacts. Quite a number of different metal contacts have been attempted, including: ohmic contacts with thick Au for better heat conduction, refractory metals such as tungsten, and variations of the ohmic metal recipe. However, the current densities in a PCSS are so high that there is no example of metals in other semiconductor technologies that will withstand these current densities.
There is an unmet need in the art for a PCSS that can survive many cycles at elevated current levels without significant damage.
SUMMARY OF THE INVENTION
The introduction of tailored doping profile (TDP) zones in the regions of the PCSS beneath and about the contacts produces a higher performance PCSS that will operate successfully in high current, high repetition applications. A semiconductor region that surrounds at least the edge of the contact is selectively doped to a depth and lateral extent such that the energy density of the PCSS filament is reduced as the current is collected at the contact. The current filament in PCSS is tens to hundreds of microns in diameter and both lateral and vertical tailoring of the doping profile is desirable. The goal is to reduce the energy density at the contact by both 1) removing energetic events from the proximity of the contact as much as possible through a reduction of the electric field near the contact so as to suppress collective impact ionization and consequent filament formation and by 2) reducing the current density at the contact through enhanced current spreading. The former is achieved by tailoring the lateral doping profile (in the direction of current flow) and the latter is achieved by tailoring the lateral and vertical (into the substrate) doping profiles. The tailored doping profile is fabricated in such a way so as not to degrade other desirable aspects of the PCSS such as high standoff voltage, resistance to surface flashover, ability to be triggered optically, and the like.
REFERENCES:
patent: 5567971 (1996-10-01), Jackson et al.
Allerman Andrew A.
Baca Albert G.
Brown Darwin J.
Denison Gary J.
Helgeson Wesley D.
Cone Gregory A.
Le Que T.
Sandia Corporation
LandOfFree
High gain photoconductive semiconductor switch having... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High gain photoconductive semiconductor switch having..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High gain photoconductive semiconductor switch having... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2443560